INDEX

Explanations

negative descriptors such as 'boring' and related terms

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to being boring or tedious

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Il

-0.74

ossession

-0.71

orns

-0.70

si

-0.68

encers

-0.67

SEA

-0.67

OURCE

-0.66

sie

-0.66

reens

-0.66

cussion

-0.66

POSITIVE LOGITS

 boring

1.11

 resil

0.99

 crap

0.88

glers

0.81

 stale

0.81

 quot

0.76

 filler

0.76

ciating

0.75

nesday

0.75

 lame

0.75

Activations Density 0.011%

negative descriptors such as 'boring' and related terms

words and phrases related to being boring or tedious

No Comments

No Known Activations